Pharmacokinetics of Direct Brain Infusion. Distribution of therapeutic agents in the central nervous system (CNS) with currently available delivery techniques is problematic. An approach that our colleagues and we developed to overcome the obstacles associated with current CNS drug delivery techniques is convective delivery, which uses bulk flow to enhance distribution. Our studies have demonstrated that convection-enhanced delivery to the brain, brainstem, spinal cord, and peripheral nerve in large and small animals can be used to distribute macromolecules in a homogenous, targeted, and safe manner and with clinically effective Vd. Recent efforts have focused on determining the factors that optimize convection-enhanced delivery into the brain, brainstem, spinal cord, and peripheral nerve. To further our understanding of the variables that affect convective delivery we are examining the effect of particle size and surface charge on distributiosn in brain, brainstem, spinal cord, and peripheral nerves. High-flow interstitial infusion is being used to deliver various agents, such as immunotoxins, recombinant enzymes, genetic vectors, and chemotherapeutic molecules in the investigation of treatment of various disorders of the central nervous system. We have developed new imaging contrast agents for non-invasively monitoring infusion volume of distribution and concentration for CT and MRI. Neurotoxicity testing of these imaging agents revealed no evidence of neuronal degeneration up to three months after infusion in rat and primate brains. Thus, we have demonstrated the utility of these imaging agents for real-time, non-invasive monitoring of the distribution of therapeutic agents during infusion of macromolecular drugs. Clinical investigations are planned including direct convective delivery of chemotherapeutic agents for treatment of brainstem gliomas, recurrent glioblastomas, and metabolic storage diseases with CNS involvement. Convection-enhanced selective excitotoxic ablation of the neurons of the globus pallidus interna (Gpi) for treatment of parkinsonism. The ability to inhibit or augment specific neuronal populations within the CNS reliably by using currently available therapeutic techniques is limited. To overcome these problems, we showed that convection-enhanced infusion of an excitotoxin can be used to selectively lesion grey matter regions of the non-human primate brain. The results suggest that this method could be used for chemical neurosurgery for medically- refractory PD and that it may be ideal for cell-specific therapeutic ablation or trophic treatment of other targeted structures associated with the CNS disorders. Treatment of Neuropathic Pain. Neuropathic pain affects millions of people and costs society tens of billions of a year in the United States alone. Recently, neuropathic pain has been found to be mediated by nociceptive neurons that selectively express the vanilloid receptor 1 (VR1). Resiniferatoxin is an excitotoxic VR1-agonist that causes destruction of VR1-positive neurons. To determine if resiniferatoxin can be used to selectively ablate VR1-positive neurons and eliminate neuropathic pain without affecting tactile sensation and motor function, we infused it using convective delivery unilaterally into primate trigeminal ganglia and tested tactile and hyperalgesic sensation in trigeminal-innervated territories The results are consistent with selective elimination of hyperalgesia on the side of resiniferatoxin-treated ganglia. Analysis revealed selective ablation of VR1-positive neurons in the resiniferatoxin-treated ganglia. Thus, VR1-positive nociceptive neurons can be safely and selectively ablated by intraganglionic resiniferatoxin infusion, resulting in elimination of hyperalgesia and neurogenic inflammation, while maintaining intact tactile sensation and motor function. The hippocampus is the usual site of origin of medically intractable epilepsy. Relief of this type of epilepsy could occur if a method were developed to selectively suppress the epileptic focus within the hippocampus. After success in ablating seizures in a rodent model using convective perfusion of the epileptic focus, we are planning a clinical study (Protocol 00-N-0158) of the infusion of muscimol into the hippocampus to temporarily inactivate the neurons of the epileptic focus in patients. If this is successful, we will explore if agents can be used to permanently and selectively inactivate the epileptic focus. As a first step in the project, we finished a study of the toxicity and distribution of muscimol into the hippocampus of non-human primates using convective delivery. Depth electrode studies showed that electrical activity in the hippocampus could be suppressed by muscimol. Autoradiography demonstrated that muscimol was distributed to the entire hippocampus using convective perfusion. The infusions were tolerated without brain injury or permanent adverse effects. The clinical protocol will be initiated after FDA aproval. In an ongoing clinical study of patients with medically intractable epilepsy (Protocol 02-N-0014), temporal lobe tissue specimens that were removed to treat medically intractable epilepsy associated with mesial temporal sclerosis were subjected to extensive histological and virological evaluation. This evaluation demonstrated that medial temporal lobe tissue from some of these patients had elevated levels of human herpesvirus type 6 (HHV-6), suggesting a role for this virus in this chronic condition. A clinical study was completed (Protocol #96-N-0093) that evaluated the functional activity and regional blood flow in cerebral cortex exposed during awake craniotomy to treat brain tumors and medically intractable epilepsy. This study demonstrated that primary brain tumors are hypoperfused compared to normal brain and that metastatic tumors are hyperperfused compared to normal brain. Brain surrounding tumors was hypoperfused due to the tumor compressing the surrounding brain. Tumor removal resulted in improved perfusion of the brain that was in proximity to the tumor.